AtomicChat/Qwen3.5-4B-DFlash-GGUF

Quantization vs z-lab/Qwen3.5-4B-DFlash. AtomicChat/Qwen3.5-4B-DFlash-GGUF is a GGUF quantization (Q8_0) of the same weights: 25 tensors match exactly and 43 within quantization error.

Tensors changed: other 0/1, Norms 0/25, MLP 0/18, Attention 0/24

On the Hub · Downloads on Model Pulse